netbox-manager 0.20251025.2

NetBox manager

netbox-manager

A Python CLI tool that automates NetBox infrastructure management through YAML-based configuration files. It streamlines the process of importing device types, module types, and managing NetBox resources using Ansible playbooks, making it easy to define and deploy complex network infrastructures with consistent configuration management.

Installation

$ pipenv shell
$ pip install netbox-manager
$ ansible-galaxy collection install -r requirements.yml

Configuration

DEVICETYPE_LIBRARY = "example/devicetypes"
IGNORE_SSL_ERRORS = true
MODULETYPE_LIBRARY = "example/moduletypes"
RESOURCES = "example/resources"
TOKEN = ""
URL = "https://XXX.netbox.regio.digital"
VARS = "example/vars"
VERBOSE = true

Usage

$ pipenv shell
$ netbox-manager --help

 Usage: netbox-manager [OPTIONS]

╭─ Options ───────────────────────────────────────────────────────────────────────────────────────────╮
│ --limit                      TEXT  Limit files by prefix [default: None]                            │
│ --skipdtl    --no-skipdtl          Skip devicetype library [default: no-skipdtl]                    │
│ --skipmtl    --no-skipmtl          Skip moduletype library [default: no-skipmtl]                    │
│ --skipres    --no-skipres          Skip resources [default: no-skipres]                             │
│ --wait       --no-wait             Wait for NetBox service [default: wait]                          │
│ --help                             Show this message and exit.                                      │
╰─────────────────────────────────────────────────────────────────────────────────────────────────────╯

Limit Functionality

The --limit option allows you to process only specific resources by their numeric prefix. The limit is applied at the top level (files and directories in the resources directory) but not to files within directories.

Examples:

# Process only resources starting with "300"
netbox-manager --limit 300

# Process only resources starting with "200"
netbox-manager --limit 200

Behavior:

• Files: --limit 300 processes 300-devices.yml but skips 200-racks.yml
• Directories: --limit 300 processes the entire 300-devices/ directory
• Directory contents: ALL files within 300-devices/ are processed, regardless of their names
• Mixed resources: Both 300-devices.yml and 300-devices/ would be processed together

Directory Structure Example:

resources/
├── 200-racks.yml              # Skipped with --limit 300
├── 300-devices.yml            # Processed with --limit 300
├── 300-networks/              # Processed with --limit 300
│   ├── cables.yml            # Processed (all files in directory)
│   ├── interfaces.yml        # Processed (all files in directory)
│   └── vlans.yml             # Processed (all files in directory)
└── 400-maintenance.yml        # Skipped with --limit 300

This approach allows you to:

• Test specific phases of your deployment (--limit 100 for infrastructure only)
• Rerun specific resource groups after changes
• Organize complex configurations into directories without losing limit functionality

Variables

netbox-manager supports both local and global variables for flexible configuration management.

Local Variables

Variables can be defined within individual resource files using the vars section:

- vars:
    site: Discworld
    location: Ankh-Morpork
    rack: "1000"
    tenant: Testbed

- device:
    name: "{{ tenant }}-switch-1"
    site: "{{ site }}"
    location: "{{ location }}"
    rack: "{{ rack }}"

Global Variables

Global variables are defined in YAML files within the VARS directory (configured in settings.toml). These variables are automatically loaded and available in all resource files.

Global Variable Files (example/vars/):

# 000-global.yml
site: Discworld
location: Ankh-Morpork
rack: "1000"
tenant: Testbed

networks:
  oob:
    vlan: 100
    subnet: "172.16.0.0/20"
  management:
    vlan: 200
    subnet: "192.168.16.0/20"

# 100-overrides.yml - loaded after 000-global.yml
networks:
  oob:
    description: "Out-of-band management network"

environment: production

Variable Precedence:

1. Global variables are loaded first (sorted by filename)
2. Later global variable files can override earlier ones
3. Local vars sections in resource files override global variables
4. Deep merging preserves nested structures

Usage in Resource Files:

# No local vars needed - uses global variables
- device:
    name: "{{ tenant }}-node-1"
    site: "{{ site }}"
    rack: "{{ rack }}"

# Local vars override global vars
- vars:
    rack: "2000"  # Overrides global rack setting

- device:
    name: "{{ tenant }}-node-2"
    site: "{{ site }}"        # Still uses global site
    rack: "{{ rack }}"        # Uses local override

Example Configuration

The example/ directory contains a complete example configuration for a testbed setup with multiple network switches and compute nodes. This configuration demonstrates all key features of netbox-manager.

Directory Structure

example/
├── devicetypes/           # Device type definitions
│   ├── Edgecore/         # Edgecore switch models
│   │   ├── 5835-54X-O-AC-F.yaml
│   │   └── 7726-32X-O-AC-F.yaml
│   └── Other/            # Generic device types
│       ├── baremetal-device.yml
│       ├── baremetal-housing.yml
│       ├── manager.yml
│       └── node.yml
├── moduletypes/          # Module definitions (empty)
├── vars/                 # Global variable files
│   ├── 000-global.yml   # Base global variables
│   └── 100-overrides.yml # Environment-specific overrides
└── resources/            # Numbered resource files and directories
    ├── 100-initialise.yml          # Base infrastructure
    ├── 200-rack-1000.yml          # Rack and device definitions
    ├── 300-testbed-manager.yml    # Manager configuration
    ├── 300-testbed-node-*.yml     # Node configurations (0-9)
    ├── 300-testbed-switch-*.yml   # Switch configurations (0-3)
    ├── 300-testbed-switch-oob.yml # Out-of-band switch
    └── 400-networks/               # Directory with network configurations
        ├── cables.yml
        ├── interfaces.yml
        └── vlans.yml

Execution Order

Files and directories are processed by their numeric prefix. Files within numbered directories are processed together in the same execution group:

1. 100-initialise.yml: Creates base infrastructure

   • Tenant: Testbed
   • Site: Discworld
   • Location: Ankh-Morpork
   • VLANs: OOB Testbed (VLAN 100)
   • IP ranges: OOB (172.16.0.0/20), Management (192.168.16.0/20), External (192.168.112.0/20)
   • IPv6 range: fda6:f659:8c2b::/48

2. 200-rack-1000.yml: Defines physical infrastructure

   • Rack "1000" with 47 rack units
   • 11 devices (1 Manager, 10 Nodes, 5 Switches) with exact rack positions

3. 300-*.yml: Detailed configuration for individual devices

   • Network interfaces and VLAN assignments
   • IP addresses and MAC addresses
   • Cable connections between devices

4. 400-networks/: Directory containing all network configuration files

   • All YAML files within this directory are processed together
   • Files are sorted alphabetically within the directory (cables.yml, interfaces.yml, vlans.yml)

Resource Organization

netbox-manager supports two organizational approaches:

File-based (traditional):

resources/
├── 100-sites.yml
├── 200-racks.yml
├── 300-devices.yml
└── 400-networks.yml

Directory-based (new):

resources/
├── 100-infrastructure/
│   ├── sites.yml
│   └── tenants.yml
├── 200-physical/
│   ├── locations.yml
│   └── racks.yml
└── 300-devices/
    ├── managers.yml
    ├── nodes.yml
    └── switches.yml

Mixed approach:

resources/
├── 100-sites.yml          # Single file
├── 200-physical/          # Directory with multiple files
│   ├── locations.yml
│   └── racks.yml
└── 300-devices.yml        # Back to single file

Device Types

Manager (manager.yml)

• 1U server with management function
• 1x 1000Base-T (management), 2x 10GBase-T, 2x 100G QSFP28

Node (node.yml)

• 1U server for Control/Compute/Storage roles
• Identical configuration to Manager

Edgecore Switches

• 7726-32X: 32-port 100G switch (Leaf/Spine)
• 5835-54X: 54-port switch for out-of-band management

Network Architecture

The example implements a typical leaf-spine architecture:

• 2x Leaf Switches (testbed-switch-0/1): Connection to compute nodes
• 2x Spine Switches (testbed-switch-2/3): Interconnect between leafs
• 1x OOB Switch (testbed-switch-oob): Out-of-band management
• 1x Manager: Orchestration and deployment
• 10x Nodes: OpenStack Control/Compute/Storage nodes

IP Addressing

• OOB Network: 172.16.0.0/20 (Out-of-band management)
• Management: 192.168.16.0/20 (OpenStack management)
• External: 192.168.112.0/20 (Public API access)
• IPv6: fda6:f659:8c2b::/48 (Management IPv6)

Documentation

• https://docs.ansible.com/ansible/latest/collections/netbox/netbox/index.html
• https://github.com/netbox-community/devicetype-library/tree/master/device-types
• https://github.com/netbox-community/devicetype-library/tree/master/module-types